Even More PM Stats Analysisby xdg (Monsignor)
|on Jan 19, 2006 at 15:21 UTC||Need Help??|
Since there seems to be persistant interest in perlmonks statistics, I thought I should post my latest quick-and-dirty analysis based on a new data set from demerphq. (And a big thank you for demerphq, too, for extracting those data!) That data set provides some insight into reputations of nodes in various categories and for various levels of monks, but bucketed in a way to maintain privacy.
Here's a table showing the average reputation per post by level for various categories. (I excluded many categories that had low sample sizes.) Of course, this is not based on any time series so it needs to be read quite literally. Monks of a current level had an average reputation of X across all their nodes in a given category (regardless of the level they were when it was originally posted.)
I've posted a 3D chart of these data for those who like a graphical perspective.
There's a strong trend evident that the higher monk levels have higher reputations per post -- even in a category like Poetry, though that has the weakest trend. There are also some interesting outliers in the Obfuscated and Tutorials categories.
One thing that can be done with per-node-reputations is to estimate the amount of XP that they generated. (The assumptions behind the estimation break down for high-reputation nodes with high node-tension, but it's the best that can be done with the data at hand.)
That affords some insight into what categories node XP came from for monks of certain levels. Not surprisingly, more experienced monks get a greater percentage of their node XP from replies to other posts (though unfortunately, the data aren't yet available to say what category the reply is in). At the highest level, XP from Meditations is a significant outlier, perhaps due to one of the three Archbishops. (Possibly Ovid, based on a today's sampling from Selected Best Nodes.)
Again I've posted some charts to illustrate it:
Code written by xdg and posted on PerlMonks is public domain. It is provided as is with no warranties, express or implied, of any kind. Posted code may not have been tested. Use of posted code is at your own risk.